Detection Method of Abnormal Data for End Users of Energy Internet

doi:10.12204/j.issn.1000-7229.2021.10.003

Abstract

Abstract:

With the continuous advancement of the energy Internet, the degree of informatization of the power system has been constantly improved and the amount of electricity data on the end-user side has been growing rapidly. The change provides a data foundation for the detection of user energy consumption based on the big data analysis technology. To deal with the high-cost and low-efficiency problem of the traditional detection model for abnormal electricity consumption patterns, a full cycle detection model of abnormal electricity consumption is proposed, which includes data cleaning, feature screening and model training. Besides, for comprehensively considering the factors affecting the abnormal electricity consumption patterns, the evaluation index system including the power consumption slope index, the line-loss index, and the warning information index is built. The data cleaning and missed value preprocess are conducted on the initial data to improve the accuracy of abnormal electricity pattern detection, and XGBoost is used for abnormal detection. Finally, a numerical case is used to verify the availability of the proposed detection method. In terms of the detection accuracy and training time, the detection performance of XGBoost algorithm is the best by comparing it with decision tree, random forest and Adaboost.

Key words: abnormal electricity consumption patterns, XGBoost, evaluation index system, detection model, energy Internet

CLC Number:

TM714

HU Yanqin, LI Haiming, LIU Nian, FU Jiekai, HUANG Tianxiang, LI Chenglin, LI Kezhou, HU Zhiqiang, FAN Zhifu, WU Xiaoke. Detection Method of Abnormal Data for End Users of Energy Internet[J]. ELECTRIC POWER CONSTRUCTION, 2021, 42(10): 19-27.

Figures/Tables 12

Fig.1 Evaluation index system of abnormal electricity consumption pattern

Table 1 Node segmentation algorithm for XGBoost

Fig.2 Flow chart of detection of abnormal electricity consumption pattern based on XGBoost

Table 2 Confusion matrix

Fig.3 The structure of distribution network

Fig.4 Iteration times - accuracy curve

Fig.5 Accuracy curve of XGBoost algorithm

Fig.6 Training effect

Fig.7 The contribution of features

Table 3 Test results before and after data cleaning

Table 4 Test results of the five algorithms

Fig.8 The ROC curve of the four algorithms

References 24

[1]	马钊, 周孝信, 尚宇炜, 等. 能源互联网概念、关键技术及发展模式探索[J]. 电网技术, 2015, 39(11): 3014-3022.
	MA Zhao, ZHOU Xiaoxin, SHANG Yuwei, et al. Exploring the concept, key technologies and development model of energy Internet[J]. Power System Technology, 2015, 39(11): 3014-3022.
[2]	肖泽青, 华昊辰, 曹军威. 人工智能在能源互联网中的应用综述[J]. 电力建设, 2019, 40(5): 63-70.
	XIAO Zeqing, HUA Haochen, CAO Junwei. Overview of the application of artificial intelligence in energy Internet[J]. Electric Power Construction, 2019, 40(5): 63-70.
[3]	余晓丹, 徐宪东, 陈硕翼, 等. 综合能源系统与能源互联网简述[J]. 电工技术学报, 2016, 31(1): 1-13.
	YU Xiaodan, XU Xiandong, CHEN Shuoyi, et al. A brief review to integrated energy system and energy Internet[J]. Transactions of China Electrotechnical Society, 2016, 31(1): 1-13.
[4]	陈启鑫, 郑可迪, 康重庆, 等. 异常用电的检测方法: 评述与展望[J]. 电力系统自动化, 2018, 42(17): 189-199.
	CHEN Qixin, ZHENG Kedi, KANG Chongqing, et al. Detection methods of abnormal electricity consumption behaviors: Review and prospect[J]. Automation of Electric Power Systems, 2018, 42(17): 189-199.
[5]	李培, 胡国平, 黄耀廉, 等. 基于集成ELM的配电网窃电嫌疑用户智能识别[J]. 科技通报, 2018, 34(12): 159-163.
	LI Pei, Hu Jianguo, HUANG Yaolian, HUANG Yaolian, et al. Intelligent identification of power-stealing users based on ELM ensembles[J]. Bulletin of Science and Technology, 2018, 34(12): 159-163.
[6]	LEITE J B, MANTOVANI J R S. Detecting and locating non-technical losses in modern distribution networks[J]. IEEE Transactions on Smart Grid, 2018, 9(2): 1023--1032. doi: 10.1109/TSG.2016.2574714 URL
[7]	李培法, 李田, 虞荻. 基于相关度的窃电用户自动识别方法[J]. 农村电气化, 2020(9): 15-17.
	LI Peifa, LI Tian, YU Di. Automatic identified method of steal electric power based on dependent degree suited to customer[J]. Rural Electrification, 2020(9): 15-17.
[8]	王昕辰, 严以臻. 基于同期线损系统的大数据分析反窃电法研究[J]. 江苏科技信息, 2019, 36(35): 47-51.
	WANG Xinchen, YAN Yizhen. Research on large data analysis anti stealing method based on line loss system in the same period[J]. Jiangsu Science & Technology Information, 2019, 36(35): 47-51.
[9]	庄池杰, 张斌, 胡军, 等. 基于无监督学习的电力用户异常用电模式检测[J]. 中国电机工程学报, 2016, 36(2): 379-387.
	ZHUANG Chijie, ZHANG Bin, HU Jun, et al. Anomaly detection for power consumption patterns based on unsupervised learning[J]. Proceedings of the CSEE, 2016, 36(2): 379-387.
[10]	林女贵, 洪兰秀, 黄道姗, 等. 基于改进深度自编码网络的异常用电行为辨识[J]. 中国电力, 2020, 53(6): 18-26.
	LIN Nvgui, HONG Lanxiu, HUANG Daoshan, et al. Abnormal electricity consumption behaviors detection based on improved deep auto-encoder[J]. Electric Power, 2020, 53(6): 18-26.
[11]	许刚, 谈元鹏, 戴腾辉. 稀疏随机森林下的用电侧异常行为模式检测[J]. 电网技术, 2017, 41(6): 1964-1973.
	XU Gang, TAN Yuanpeng, DAI Tenghui. Sparse random forest based abnormal behavior pattern detection of electric power user side[J]. Power System Technology, 2017, 41(6): 1964-1973.
[12]	SUN Xiaoyu, SUN Xiangyang, ZHAO Yang. Abnormal user identification based on XGBoost algorithm.[J]. Journal of Measurement Science and Instrumentation, 2018, 9(4):339-346.
[13]	孙峰, 王刚, 程绪可, 等. 一种基于均值漂移和XGBoost的异常用电判别方法: CN111160401A[P]. 2020-05-15.
[14]	JINDAL A, DUA A, KAUR K, et al. Decision tree and SVM-based data analytics for theft detection in smart grid[J]. IEEE Transactions on Industrial Informatics, 2016, 12(3): 1005-1016. doi: 10.1109/TII.2016.2543145 URL
[15]	陈海焱, 陈金富, 杨雄平, 等. 配电网中计及短路电流约束的分布式发电规划[J]. 电力系统自动化, 2006, 30(21): 16-21.
	CHEN Haiyan, CHEN Jinfu, YANG Xiongping, et al. Planning for distributed generation in distribution network with short-circuit current constraints[J]. Automation of Electric Power Systems, 2006, 30(21): 16-21.
[16]	熊霞, 陶晓峰, 叶方彬, 等. 基于台区识别和关联监测加权算法的窃电检测方法[J]. 计算机应用, 2019, 39(S2): 289-292.
	XIONG Xia, TAO Xiaofeng, YE Fangbin, et al. Electricity stealing detection method based on weighted algorithm of station identification and correlation monitoring[J]. Journal of Computer Applications, 2019, 39(S2): 289-292.
[17]	ZHENG Z B, YANG Y T, NIU X D, et al. Wide and deep convolutional neural networks for electricity-theft detection to secure smart grids[J]. IEEE Transactions on Industrial Informatics, 2018, 14(4): 1606-1615. doi: 10.1109/TII.2017.2785963 URL
[18]	胡天宇, 郭庆来, 孙宏斌. 基于堆叠去相关自编码器和支持向量机的窃电检测[J]. 电力系统自动化, 2019, 43(1): 119-125.
	HU Tianyu, GUO Qinglai, SUN Hongbin. Nontechnical loss detection based on stacked uncorrelating autoencoder and support vector machine[J]. Automation of Electric Power Systems, 2019, 43(1): 119-125.
[19]	BALLAL M S, SURYAWANSHI H, MISHRA M K, et al. Online electricity theft detection and prevention scheme for smart cities[J]. IET Smart Cities, 2020, 2(3): 155-164. doi: 10.1049/iet-smc.2020.0045 URL
[20]	郑建宁. 基于深度学习的窃电行为检测方法[J]. 信息技术, 2019, 43(2): 156-159.
	ZHENG Jianning. Detection method of electricity stealing behavior based on deep learning[J]. Information Technology, 2019, 43(2): 156-159.
[21]	李庆扬. 数值分析[M]. 4版. 北京: 清华大学出版社, 2001.
[22]	李晓菲. 数据预处理算法的研究与应用[D]. 成都: 西南交通大学, 2006.
	LI Xiaofei. Research and application on data preprocessing algorithms[D]. Chengdu: Southwest Jiaotong University, 2006.
[23]	杨贵军, 徐雪, 赵富强. 基于XGBoost算法的用户评分预测模型及应用[J]. 数据分析与知识发现, 2019, 3(1): 118-126.
	YANG Guijun, XU Xue, ZHAO Fuqiang. Predicting user ratings with XGBoost algorithm[J]. Data Analysis and Knowledge Discovery, 2019, 3(1): 118-126.
[24]	张良均. Python数据分析与挖掘实战[M]. 北京: 机械工业出版社, 2015.